读取非定界 asciif 文件 Apache Pig Latin

Question

我正在尝试读取 Apache Pig Latin 中的文本文件，该文件的每一行都包含非分隔的 ascii。也就是说，该行中的每一列都在行中的特定位置开始和结束。

示例定义：

+--------+----------------+--------------+
| Column | Start Position | End Position |
+--------+----------------+--------------+
| A      | 1              | 6            |
+--------+----------------+--------------+
| B      | 8              | 11           |
+--------+----------------+--------------+
| C      | 13             | 15           |
+--------+----------------+--------------+

示例数据：

+---+---+---+---+---+---+---+----+---+----+----+----+----+----+----+
| 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8  | 9 | 10 | 11 | 12 | 13 | 14 | 15 |
+---+---+---+---+---+---+---+----+---+----+----+----+----+----+----+
| s | a | m | p | l | e |   | d  | a | t  | a  |    |    | h  | i  |
+---+---+---+---+---+---+---+----+---+----+----+----+----+----+----+
| d | u | d | e |   |   |   | hi |   |    |    |    | b  | r  | o  |
+---+---+---+---+---+---+---+----+---+----+----+----+----+----+----+

预期输出：

sample, data, hi
dude, hi, bro

如何在 Pig 中阅读此内容？ PigStorage 似乎不够灵活，无法允许位置分隔，只能使用字符串分隔（逗号、制表符等）。

Answer 1

看起来 Apache 为这个特定用例提供了一个加载器：

LOAD 'data.txt' USING org.apache.pig.piggybank.storage.FixedWidthLoader('1-6, 8-11, 13-15', 'SKIP_HEADER') AS (a, b, c);

https://pig.apache.org/docs/r0.16.0/api/

读取非定界 asciif 文件 Apache Pig Latin

Read non delimited asciif file Apache Pig Latin

parsing

apache-pig

bigdata

hadoop2